Riggers, SAR, Engineers: Help answer the hardest question on MP ever!

Okay, maybe an exaggeration, but this issue has me a bit stumped. Here is the deal. Recently I have been conducting efficiency tests on various 5:1 pulley configurations consisting of two double pulleys. To calculate the system's efficiency I threw a dyno on the output of the MA system and the input, or "pull strand," and then calculated the efficiency with the values logged by the dyno after applying 100 lbf to the input of the MA system. I used two trees about 20' apart to anchor both sides of the MA system.

Here is the problem. The very first time I conduct the test I observe accurate and otherwise repeatable results. However, if I repeat the test many times in a row in short succession (say 20 seconds apart) the output load, and accordingly the system efficiency, drops with every subsequent test even though the input load does not change.

I tried to isolate the problem by testing the efficiency of each sheave in a 1:1. This resulted in no change in efficiency after subsequent retests. Every trial for every sheave gave me close to the same efficiency value no matter how many times I repeated the test.

Next, I considered that the elongation properties of the rope may be affecting the values and so I substituted my 3/8" polyester static for a 9mm nylon low stretch. Again, same issue although the MA system's efficiency did not decline repeated testing as fast as it did with the polyester static rope. Last, I tried using dynamic rope. Again, for the first test the MA system's efficiency value was identical to the first test of other series. However, like the other trials, the efficiency of the MA system also declined with subsequent retests, but like the low-stretch rope, the decline was less steep with the dynamic rope than it was with the static. Last, I tried swapping the pulleys with other double pulleys. No change, same issue.

I have spoken with at least one engineer from three of the largest pulley manufacturers in the industry and they have no idea or theory as to what is causing the issue. In addition, because I know it will come up, I am positive the dynos are accurate and the scan rate is sufficiently high enough to capture the peak load.

Any ideas?

Caveat #1 - I have no engineering experience with pulleys, other than sizing them for various applications to get the ratio needed.

Caveat #2 - Some of the stuff I am going to say may be obvious, but I've always tried to follow the KISS principle.

You already see a difference with the rope, which was my first guess.

That said, gut feeling at 7 a.m. with only one cup of coffee is to next look at the heat and friction. If you are doing repeated tests, how much time are you allowing the system to come back to steady state? Any system with moving parts doing work is going create heat from friction which will be seen as efficiency loss.

Just my $.02 so take with a grain of salt.

Cool test, BTW.

A picture of your setup might be useful.

My first thought is the rope.

If you could get a hold of some steel cable (even small diameter stuff) and run several tests in a row I would guess that you'd see consistent results.

I doubt that the heat produced in the pulleys would affect the efficiency in any appreciable way.

I'm curious to see what others postulate.

Probably what you were trying to head off but is it possible that the repeated cycles arent allowing your force measuring devices to return to their "zero- state" completely and your getting creep in your zero. I would be surprised if there was anything mechanical besides the rope.

My thought is since the rope isn't really flowing thru the system, but being pulled to the same spot every time, the yarns within the rope are kind of taking a "set" around the pulley, which is acting as a hidden source of drag.

For your 5:1, is there a fixed end of the rope in the pulley system? Maybe you could fix that end w/ a rope clamp rather than a knot, and advance some fresh rope into the system quickly (even just shuffling it a foot or so), so there's a new area of rope on the sheaves.

maybe try cable as a control

Thoughts on avg contact pressure w/ the pulleys? -- psi sheave pressure from mainlines stressing to the pulley groove, along with the coef of friction that goes along with the pulley. Increasing the pulley/sheave diameter might help.

other thought is the z axis?

Further:
I had some points on the original slackline setup, but can't recall if you displayed all the components in an itemized fashion with equipment specs. The main problem I had was the presentation format, so it would be hard to peer review. Granted, it's online discussion, it's quick & easy to put up pics and discussion to shape the experiment.

Hey 4.5 kips,
I meet all 3 of your criteria. I wonder how you are applying the input load and if dynamics are involved...

bearing friction is a function of temperature.

When you say you repeat the tests in quick succession, does that mean you release the 100lbf load and let the entire system go slack or is there still a small amount of tension in the system in between each quick test?

A picture of the exact type of pulleys might be helpful as well.

rope deformation. sounds like the cords with a sheath are working better than the polyester cord. Maybe the cord if flattening and hitting the sides of the sheaves?.

Larry S wrote:My thought is since the rope isn't really flowing thru the system, but being pulled to the same spot every time, the yarns within the rope are kind of taking a "set" around the pulley, which is acting as a hidden source of drag. For your 5:1, is there a fixed end of the rope in the pulley system? Maybe you could fix that end w/ a rope clamp rather than a knot, and advance some fresh rope into the system quickly (even just shuffling it a foot or so), so there's a new area of rope on the sheaves.

I think Larry is the winner. In a static system the elasticity of the rope should not matter, so its something else about the different ropes that makes the difference. I agree with Larry that it's likely the stiffness of the rope.

good question. several people have mentioned some good ideas:

load cell/data acquisition system no fully returning to zero, causing a floating zero.

creep in the rope

deviations in contact pressure on the pulleys.

another thing i would look at - are you measuring the deflection of your anchor (ie the trees). in particular, is the frequency of your repetions close to the natural frequency of the bending mode of the trees? this would probably cause what you are describing. also, if there is much repetitive deflection in the tree, you could be getting some creep there, or some other effects on the bending stiffness(?).

good luck and keep us posted!

Pitch it to the Rope Test Lab guys... They do lots of stuff like this.

I like the rope set (Larry) theories. Explains the differing rates bewteen rope types.

You don't say what your results are but I'm gonna guess around 2.6:1 actual or less on the first pull?

This question was actually in reference to an event that occurred months ago, I just thought of it now so I posted it. But with all the response I decided to see if I could replicate the results I was experiencing a few months ago. I couldn't replicate the results from the last test. This time the efficiency value was very repeatable. Although it is worthwhile to note I am using a different set of pulleys now than I did when I first observed the issue months ago. Oh well, it's a shoulder shrugger.

The orange represents the load on the cell on the output-side of the MA system and the blue is the input. The two plots dont match, in time, on the graph, but only because I did not take the time to correct it on the graphs. In the real testing, the peaks of each plot match.



As far as cable, the pulley uses AL sheaves, so no steel cable. As far as the zero balance error goes, the official rating is 1% FS maximum, but that is a severe overrating, the real value is more like <0.1%. The load cells zero out quite well. I have pulled one of my cells to its maximum safe readable value (5,000 lbs) and the cell returned to within one pound of true zero. The worst drift I have ever seen was 2.5 pounds and that was on my 10,000 lb cell with the A/D converter scanning at maximum speed.

20 kN wrote: As far as cable, the pulley uses AL sheaves, so no steel cable.

Any of the tech fiber 12-strand ropes would work just fine for this. They've gotten quite popular with the logging folks and the 4wd crowd since they are just as strong for size, don't rust, are lighter, and don't get needles sticking out of them as the cable wears. Easier to splice too. Easy to get from a yachting supplier. I personally prefer dyneema/vectran blend (e.g. Samson Lightning) as it's a bit less slippery, but the full dyneema ones (e.g. Samson Amsteel) work just as well. NE Ropes makes some nice ones too, BTW.

20 kN wrote: I couldn't replicate the results from the last test. This time the efficiency value was very repeatable. Although it is worthwhile to note I am using a different set of pulleys now than I did when I first observed the issue months ago.

I thought it had something to do with the pulleys, not the rope. I can't quite imagine what the problem was with the pulleys, but as far as the rope is concerned, it seems to me that rope set (i.e. the rope stretching and not returning to original length/shape) should increase the load transfer, not reduce, as there would be overall less 'play' in the system.

Without a control, it's difficult to isolate the behaviors. I like Aric's idea of getting some dyneema/tech cord and going that route. I would incorporate more rescue-type larger diameter pulleys, also.

20kN - Your rig looks like a dedicated/designed 5:1 "block and Tackle" system, not a field-rigged system. Still that is some surprising efficiency. Any datasets with 3:1 or 2x3:1's using typical gear (biners or small pulleys) of a field rigged emergency haul system?? No climber is apt to have the rig in your pic....just curious

I typically do a Z with my Reverso4 in autolock mode at the point and add a 2:1 if needed. Simple, fast, little/no extra gear, takes the efficiency hit (luckily my pipes can make up the deficit) but curious what your load cells would say.

I would go to Rigging For Rescue or CMC for your answers. They are THE experts!

Matt